VERNERSCHOMAKER AND ROBERT SPURR
11%
correspond to them in the spectrum of acetylmesitylene. Again, the relatively high value of 1699 cm.-l for the C=O frequency indicates that the ring and carbonyl group are not coplanar.
Summary Raman frequencies, estimated intensities and depolarization factors are reported for acetophenone, mesitylaldehyde, acetylmesitylene, methyl
[CONTRIBUTION PROM THE
Vol. 64
2,4,6-trimethylbenzoate and 2,4,6-trimethylbenzoyl chloride; also, frequencies and intensities are reported for acetyldurene and nitromesitylene. For acetophenone, the only one for which previous data seem to have been reported, new lines were observed at 404, 896 and 3006 em.-' and depolarization data were obtained for eleven additional lines. CHICAGO, ILLINOIS
RECEIVED NOVEMBER 3, 1941
GATESAND CRELLIN LABORATORIES OF CHEMISTRY, CALIFORNIA INSTITUTE OF TECHNOLOGY No. $851
The Structures of Nitrous Oxide and of Hydrogen Azide BY VERNERSCHOMAKER AND ROBERT SPURR Introduction.-Nitrous oxide and hydrogen azide present much the same structural problem. The two molecules contain the same number of valence electrons and have been shown by spectroscopic investigation1i15to be essentially linear. The expected resonating structures are similar ; three which might be considered important are
-
(I)
i
:N=N=O:
-
-+
:N = N =
H 6
7
8
10
k2, we find that 71 = 1.14 A. and 7 2 = 1.21 A,, which give the ratio 0.94, in agreement with the distances predicted for resonance between structures I 11. Using the same interaction force constant and the parallel frequencies for hydrogen azide,l1 we find that r3 = 1.14 A. and r4 = 1.25 A.,in excellent agreement with the electron diffraction values. The distance rl 7 2 = 2.32 * 0.02 A. agrees
+
+
(10) A. Adel and D. Dennison, Phys. Rev., 48, 722 (1933). (11) E.H. Eyster, J . Chcm. P k y s . , 8, 373 (1940)
May, 1942
iis?
REDUCTION OF 6OMPLkX NICKEL CYANIbkS
with the determinations by electron diffraction of WierP (2.38 * 0.08 Hi.) and Maxwell, Mosley, and Demingla (2.38 * 0.05 A.), and with the length predicted from the moment of inertial4 (2.31 k).The structure found for hydrogen azide is identical with that proposed by Eyster’5 from a consideration of the moments of inertia, and on the assumption that the resonating structures are chiefly I 11.
azide have been investigated by the electron diffraction method. For nitrous oxide rl YZ = 2.32 * 0.02 b.,and 0.925 < r1/r2 < 1.08. The distances y1 = 1.12 b. and rz = 1.19 A. predicted by Pauling on the basis of his adjacent charge rule give a ratio, 0.94, which lies between these limits. It is shown that the distances found by the use of Badger’s rule also have the ratio 0.94 For hydrogen azide the distances found are YQ = 1.136 * 0.01 A. and r4 = 1.247 * 0.01 A., on the basis of an assumed linear N3 group. The report of Eyster, made on the basis of his values of the moments of inertia and Pauling’s adjacent charge rule, is confirmed. I t is shown that the application of Badger’s rule yields similar results.
+
+
summary The structures of nitrous oxide and hydrogen (12) R . Wierl, Physik. Z . , 81, 1028 (1930). (13) L. Maxwell, V. Mosley and L. Deming, J . Chcm. Phys., 2, 331 (1934). (14) E. F. Barker, Phys. Rev., 41, 369 (1932). (15) E. H. Eyster, J . Chcm. Phys., 8 , 141 (1949).
PASADENA, CALIFORNIA
RECEIVED JANUARY 7, 1942
[CONTRIBUTION FROM THE DEPARTMENT OF CHEWISTRY, UNIVERSITY OF CINCINNATI]
A Study of the Products Obtained by the Reducing Action of Metals upon Salts in Liquid Ammonia Solutions. VII. The Reduction of Complex Nickel Cyanides: Mono-valent Nickel BY JOHN W. EASTES AND WAYLAND M. BURGESS’ The red color produced by the reduction of aqueous solutions of potassium cyanonickelate, &Ni(CN)*, was shown by Bellucci and Corelli2 to be attributable t o a substance having the empirical formula &Ni(CNh. This substance had, however, never been isolated. In the course of a study of the reduction of nickel salts in anhydrous liquid ammonia,a KtNi(CN)a was observed to be formed and we have now succeeded in isolating it in a pure condition, both from solutions of liquid ammonia and from aqueous solutions, and subjecting it to a complete analysis and study. Materials Used Alkali Metals.-Kahlbaum metallic sodium and potassium were used. The sodium was analyzed by solution in 95% ethanol, conversion to sodium chloride by the addition of excess hydrochloric acid, evaporation t o dryness and weighing as sodium chloride. An average of four analyses gave a value of 100.5% sodium, with a maximum deviation of 0.2%. The potassium was analyzed in a (1) This article is based upon the thesis presented to the Faculty of the Graduate School of the University of Cincinnati by John W. Eastes in partial fulfillment of the requirements for the degree of Doctor of Philosophy, 1936. No VI, THIS J O U R N A L , 88, 2674 (1941)
(2) Bellucci and Corelli, Z anorg. Chrm., 86, 88-104 (1914). (3) Burgess and Easte- TRISJOURNAL, 68, 2674-6 (1941).
.
similar manner, but with the addition that it was also precipitated and weighed as K Z N ~ C O ( N O ~ ) . ~and H~O~ found t o be 100.4% potassium, with a maximum deviation of o.4%o. sodium and Potassium Cyanonickelate; NazNi(CN),, K2Ni(CN),.-These two compounds were prepared in a pure form by recrystallization of the product obtained by concentrating a dilute solution formed by dissolving freshly precipitated nickel cyanide in a solution of the alkali metal cyanide. The anhydrous products had the analyses: for Na2Ni(CN)4-28.07% Ni (calcd. 28.12% Ni), 49.93% CN (calcd. 49.85% CN); for KZNi(CN)4-.24.30% Ni (calcd. 24.36% Ni), 43.56% CN (calcd. 43.18% CN). Preparation of KzNi( CN)3 in Liquid Ammonia.-Reactions in anhydrous liquid ammonia were carried out as described by Burgess and Eastes,8 a diagram of the apparatus used being shown in Fig. 1. By the reaction of excess potassium cyanonickelate with sodium or potassium, K2Ni(CN)awas prepared in liquid ammonia solution as follows. The reaction tube was twothirds filled with anhydrous liquid ammonia, which was then saturated with potassium cyanonickelate. To this solution was added a small piece, approximately 0.1 g., of sodium or potassium. A bright red, dense precipitate slowly formed in the solution and settled out a t once. No hydrogen was formed during the reaction. When the alkali metal had all reacted, the yellow solution above th? red precipitate was siphoned off and the red precipitatc. (4) Hamid, Analyst. 61, 460 (1926).
